© Neuronpedia 2026
    Privacy & TermsBlogGitHubSlackTwitterContact
    Neuronpedia logo - a computer chip with a rounded viewfinder border around it

    Neuronpedia

    APIAssistant AxisNEWCircuit TracerNEWSteerSAE EvalsExports Community BlogPrivacy & TermsContact
    1. Home
    3. Joseph Bloom · Feature Splitting for GPT2-Small
    5. GPT2-Small
    7. Residual Feature Splitting 24576
    9. 8-RES_FS24576-JB
    11. 9329
    Prev
    Next
    INDEX
    Explanations

    technical terms related to networks and signals, especially focusing on signaling pathways and network architecture

    oai_token-act-pair · gpt-3.5-turboTriggered by @bot
    New Auto-Interp
    Top Features by Cosine Similarity
    Comparing With GPT2-SMALL @ 8-res_fs24576-jb
    Configuration
    jbloom/GPT2-Small-Feature-Splitting-Experiment-Layer-8/blocks.8.hook_resid_pre_24576
    Prompts (Dashboard)
    24,576 prompts, 128 tokens each
    Dataset (Dashboard)
    Skylion007/openwebtext
    Features
    24,576
    Data Type
    torch.float32
    Hook Point
    blocks.8.hook_resid_pre
    Architecture
    standard
    Context Size
    128
    Dataset
    Skylion007/openwebtext
    Hook Point Layer
    8
    Activation Function
    relu
    Embeds
    IFrame
    Link
    Not in Any Lists

    No Comments

    Neuron Alignment
    Index
    Value
    % of L₁
    288
    +0.17
    0.8%
    649
    +0.11
    0.5%
    447
    +0.11
    0.5%
    Correlated Neurons
    Index
    P. Corr.
    Cos Sim.
    288
    +0.17
    0.03
    41
    +0.11
    0.02
    325
    +0.11
    0.03
    Negative Logits
    ilts
    -0.68
    erenn
    -0.67
    OPLE
    -0.67
    ugh
    -0.67
    orical
    -0.67
    DonaldTrump
    -0.66
    mond
    -0.66
     Lama
    -0.64
     Fraz
    -0.64
    gg
    -0.61
    POSITIVE LOGITS
     networks
    1.00
     connectivity
    0.98
     network
    0.92
     routers
    0.91
     architectures
    0.89
     connection
    0.85
     connections
    0.84
     Networks
    0.83
    ed
    0.81
     infrastructure
    0.81
    Activations Density 0.036%

    No Known Activations